Relationship Between Hyperglycemia and Cognitive Function in Older NIDDM Patients
The nature and extent of cognitive impairment was examined in 29 healthy elderly subjects (mean age 69.8 yr) with non-insulin-dependent diabetes mellitus (NIDDM) and 30 demographically similar nondiabetic community volunteers (mean age 68 yr). Measures of verbal learning, abstract reasoning, and complex psychomotor functioning were performed more poorly by diabetic than nondiabetic subjects. Conversely, there were no between-group differences in performance on tasks involving pure motor speed and simple verbal abilities. Within the diabetic group, individuals with poorer metabolic control performed more poorly on tasks involving learning, reasoning, and complex psychomotor performance, although this relationship was not evident for simple verbal or motor tasks. These data indicate that older people with NIDDM who are functioning well and perceive themselves as in good health are likely to manifest greater deficits than healthy elderly people in processing complex verbal or nonverbal material. Possible explanatory mechanisms are discussed, and directions for future research are explored. Diabetes Care 13:16-21, 1990
lthough the prevalence of non-insulin-dependent diabetes mellitus (NIDDM) increases with age, neither the clinical impact of this phenomenon nor the therapeutic approach to this patient population has been well defined. There are many
A
From the Department of Medicine, Stanford University School of Medicine and Geriatric Research, Education and Clinical Center, Veterans Administration Medical Center, Palo Alto, California. Address correspondence and reprint requests to Gerald M. Reaven, MD, GRECC (182-B), Veterans Administration Medical Center, 3801 Miranda Avenue, Palo Alto, CA 94304. Received for publication 18 August 1988 and accepted in revised form 17 July 1989.
16
Gerald M. Reaven, MD Larry W. Thompson, PhD David Nahum, BS Edmund Haskins, PhD
reasons for this state of affairs. For example, efforts to prevent development of the vascular complications associated with diabetes tend to be downplayed in older patients, presumably because the average life expectancy in this population does not justify the endeavor. This consideration, in juxtaposition with concern over the harmful impact of hypoglycemia in older individuals, leads to a treatment approach that often emphasizes doing no harm at the expense of achieving good glycemic control. However, recognition of the dangers of hypoglycemia should not obviate consideration of the possibility that poor glycemic control may also have untoward effects in older patients with NIDDM. In particular, we were concerned with earlier studies that indicated some measurements of cognitive function were reduced in elderly patients with NIDDM compared with control subjects of similar age (1-5). On the other hand, other studies failed to detect differences in cognitive function between healthy control subjects and patients with NIDDM who were otherwise in good health (6,7). Although these conflicting results await a complete explanation, it has been suggested that factors other than diabetes may be responsible. For example, Mattlar et al. (6) raised the possibility that failure to find a difference between diabetic patients and control subjects might be because they had excluded all subjects with cardiovascular diseases (CVD). Given this situation, it seemed reasonable to initiate this study, which had the following two goals. First, we compared multiple measures of cognitive function in a population of patients with NIDDM >60 yr of age and a group of nondiabetic subjects matched for age and other relevant variables. In particular, we used measures covering a wide range of abilities of varying levels of difficulty to determine the intertest pattern and scope of
DIABETES CARE, VOL. 13, NO. 1, JANUARY 1990
C M . REAVEN AND ASSOCIATES
group differences. Second, we wanted to determine whether there was any correlation between the measures of cognitive function used and degree of glycemic control in diabetic patients.
RESEARCH DESIGN AND METHODS Patients with NIDDM (n = 29) and nondiabetic subjects (n = 30) were selected from a larger group of subjects who were participating in a longitudinal study evaluating the impact of NIDDM on physical and psychological functioning in the elderly. The nondiabetic subjects were obtained from a panel of community elders who volunteered to participate in geriatric research at our facility. They were determined to be nondiabetic on the basis of history and physical examination and standard laboratory values for evaluating fasting plasma glucose (FPG) and glycosylated hemoglobin (HbA1c). Patients were considered to have NIDDM if that diagnosis had been applied to them in the past and they were being treated with diet alone or diet plus sulfonylurea compounds. None of the patients gave a history of ketonuria or diabetic ketoacidosis. Urinalyses for ketones were negative when measured during the hospital stay. These characteristics make it probable that all or most of the diabetic subjects had NIDDM. An attempt was made to select both diabetic and control subjects with comparable sociodemographic characteristics, who were highly functional in terms of occupational history and current daily activities, and who perceived themselves to be in reasonably good physical and mental health. Initial criteria for selection of both groups included 7) >59 yr of age; 2) >2 yr of high school; 3) occupation above the level of unskilled laborer; 4)
The nature and extent of cognitive impairment was examined in 29 healthy elderly subjects (mean age 69.8 yr) with non-insulin-dependent diabetes mellitus (NIDDM) and 30 demographically similar nondiabetic community volunteers (mean age 68 yr). Measures of verbal learning, abstract reasoning, and complex psychomotor functioning were performed more poorly by diabetic than nondiabetic subjects. Conversely, there were no between-group differences in performance on tasks involving pure motor speed and simple verbal abilities. Within the diabetic group, individuals with poorer metabolic control performed more poorly on tasks involving learning, reasoning, and complex psychomotor performance, although this relationship was not evident for simple verbal or motor tasks. These data indicate that older people with NIDDM who are functioning well and perceive themselves as in good health are likely to manifest greater deficits than healthy elderly people in processing complex verbal or nonverbal material. Possible explanatory mechanisms are discussed, and directions for future research are explored. Diabetes Care 13:16-21, 1990
lthough the prevalence of non-insulin-dependent diabetes mellitus (NIDDM) increases with age, neither the clinical impact of this phenomenon nor the therapeutic approach to this patient population has been well defined. There are many
A
From the Department of Medicine, Stanford University School of Medicine and Geriatric Research, Education and Clinical Center, Veterans Administration Medical Center, Palo Alto, California. Address correspondence and reprint requests to Gerald M. Reaven, MD, GRECC (182-B), Veterans Administration Medical Center, 3801 Miranda Avenue, Palo Alto, CA 94304. Received for publication 18 August 1988 and accepted in revised form 17 July 1989.
16
Gerald M. Reaven, MD Larry W. Thompson, PhD David Nahum, BS Edmund Haskins, PhD
reasons for this state of affairs. For example, efforts to prevent development of the vascular complications associated with diabetes tend to be downplayed in older patients, presumably because the average life expectancy in this population does not justify the endeavor. This consideration, in juxtaposition with concern over the harmful impact of hypoglycemia in older individuals, leads to a treatment approach that often emphasizes doing no harm at the expense of achieving good glycemic control. However, recognition of the dangers of hypoglycemia should not obviate consideration of the possibility that poor glycemic control may also have untoward effects in older patients with NIDDM. In particular, we were concerned with earlier studies that indicated some measurements of cognitive function were reduced in elderly patients with NIDDM compared with control subjects of similar age (1-5). On the other hand, other studies failed to detect differences in cognitive function between healthy control subjects and patients with NIDDM who were otherwise in good health (6,7). Although these conflicting results await a complete explanation, it has been suggested that factors other than diabetes may be responsible. For example, Mattlar et al. (6) raised the possibility that failure to find a difference between diabetic patients and control subjects might be because they had excluded all subjects with cardiovascular diseases (CVD). Given this situation, it seemed reasonable to initiate this study, which had the following two goals. First, we compared multiple measures of cognitive function in a population of patients with NIDDM >60 yr of age and a group of nondiabetic subjects matched for age and other relevant variables. In particular, we used measures covering a wide range of abilities of varying levels of difficulty to determine the intertest pattern and scope of
DIABETES CARE, VOL. 13, NO. 1, JANUARY 1990
C M . REAVEN AND ASSOCIATES
group differences. Second, we wanted to determine whether there was any correlation between the measures of cognitive function used and degree of glycemic control in diabetic patients.
RESEARCH DESIGN AND METHODS Patients with NIDDM (n = 29) and nondiabetic subjects (n = 30) were selected from a larger group of subjects who were participating in a longitudinal study evaluating the impact of NIDDM on physical and psychological functioning in the elderly. The nondiabetic subjects were obtained from a panel of community elders who volunteered to participate in geriatric research at our facility. They were determined to be nondiabetic on the basis of history and physical examination and standard laboratory values for evaluating fasting plasma glucose (FPG) and glycosylated hemoglobin (HbA1c). Patients were considered to have NIDDM if that diagnosis had been applied to them in the past and they were being treated with diet alone or diet plus sulfonylurea compounds. None of the patients gave a history of ketonuria or diabetic ketoacidosis. Urinalyses for ketones were negative when measured during the hospital stay. These characteristics make it probable that all or most of the diabetic subjects had NIDDM. An attempt was made to select both diabetic and control subjects with comparable sociodemographic characteristics, who were highly functional in terms of occupational history and current daily activities, and who perceived themselves to be in reasonably good physical and mental health. Initial criteria for selection of both groups included 7) >59 yr of age; 2) >2 yr of high school; 3) occupation above the level of unskilled laborer; 4)
